src/HOL/Auth/OtwayRees_Bad.ML
author paulson
Fri, 17 Jan 1997 12:49:31 +0100
changeset 2516 4d68fbe6378b
parent 2451 ce85a2aafc7a
child 2637 e9b203f854ae
permissions -rw-r--r--
Now with Andy Gordon's treatment of freshness to replace newN/K
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
     1
(*  Title:      HOL/Auth/OtwayRees_Bad
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
     2
    ID:         $Id$
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
     3
    Author:     Lawrence C Paulson, Cambridge University Computer Laboratory
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
     4
    Copyright   1996  University of Cambridge
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
     5
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
     6
Inductive relation "otway" for the Otway-Rees protocol.
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
     7
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
     8
The FAULTY version omitting encryption of Nonce NB, as suggested on page 247 of
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
     9
  Burrows, Abadi and Needham.  A Logic of Authentication.
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    10
  Proc. Royal Soc. 426 (1989)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    11
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    12
This file illustrates the consequences of such errors.  We can still prove
2032
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
    13
impressive-looking properties such as Spy_not_see_encrypted_key, yet the
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    14
protocol is open to a middleperson attack.  Attempting to prove some key lemmas
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    15
indicates the possibility of this attack.
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    16
*)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    17
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    18
open OtwayRees_Bad;
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    19
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    20
proof_timing:=true;
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    21
HOL_quantifiers := false;
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    22
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    23
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    24
(*Weak liveness: there are traces that reach the end*)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    25
goal thy 
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    26
 "!!A B. [| A ~= B; A ~= Server; B ~= Server |]   \
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    27
\        ==> EX K. EX NA. EX evs: otway.          \
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
    28
\               Says B A {|Nonce NA, Crypt (shrK A) {|Nonce NA, Key K|}|} \
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    29
\                 : set_of_list evs";
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    30
by (REPEAT (resolve_tac [exI,bexI] 1));
2032
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
    31
by (rtac (otway.Nil RS otway.OR1 RS otway.OR2 RS otway.OR3 RS otway.OR4) 2);
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
    32
by possibility_tac;
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    33
result();
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    34
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    35
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    36
(**** Inductive proofs about otway ****)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    37
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    38
(*Nobody sends themselves messages*)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    39
goal thy "!!evs. evs : otway ==> ALL A X. Says A A X ~: set_of_list evs";
2032
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
    40
by (etac otway.induct 1);
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    41
by (Auto_tac());
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    42
qed_spec_mp "not_Says_to_self";
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    43
Addsimps [not_Says_to_self];
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    44
AddSEs   [not_Says_to_self RSN (2, rev_notE)];
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    45
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    46
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    47
(** For reasoning about the encrypted portion of messages **)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    48
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    49
goal thy "!!evs. Says A' B {|N, Agent A, Agent B, X|} : set_of_list evs ==> \
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
    50
\                X : analz (sees lost Spy evs)";
2032
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
    51
by (fast_tac (!claset addSDs [Says_imp_sees_Spy RS analz.Inj]) 1);
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
    52
qed "OR2_analz_sees_Spy";
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    53
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
    54
goal thy "!!evs. Says S B {|N, X, Crypt (shrK B) X'|} : set_of_list evs ==> \
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
    55
\                X : analz (sees lost Spy evs)";
2032
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
    56
by (fast_tac (!claset addSDs [Says_imp_sees_Spy RS analz.Inj]) 1);
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
    57
qed "OR4_analz_sees_Spy";
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    58
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
    59
goal thy "!!evs. Says Server B {|NA, X, Crypt K' {|NB,K|}|} : set_of_list evs \
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    60
\                 ==> K : parts (sees lost Spy evs)";
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
    61
by (fast_tac (!claset addSEs sees_Spy_partsEs) 1);
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    62
qed "Oops_parts_sees_Spy";
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    63
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    64
(*OR2_analz... and OR4_analz... let us treat those cases using the same 
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    65
  argument as for the Fake case.  This is possible for most, but not all,
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    66
  proofs: Fake does not invent new nonces (as in OR2), and of course Fake
2032
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
    67
  messages originate from the Spy. *)
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    68
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
    69
bind_thm ("OR2_parts_sees_Spy",
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
    70
          OR2_analz_sees_Spy RS (impOfSubs analz_subset_parts));
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
    71
bind_thm ("OR4_parts_sees_Spy",
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
    72
          OR4_analz_sees_Spy RS (impOfSubs analz_subset_parts));
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
    73
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    74
val parts_Fake_tac = 
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
    75
    forward_tac [OR2_parts_sees_Spy] 4 THEN 
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
    76
    forward_tac [OR4_parts_sees_Spy] 6 THEN
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    77
    forward_tac [Oops_parts_sees_Spy] 7;
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    78
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    79
(*For proving the easier theorems about X ~: parts (sees lost Spy evs) *)
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    80
fun parts_induct_tac i = SELECT_GOAL
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    81
    (DETERM (etac otway.induct 1 THEN parts_Fake_tac THEN
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
    82
             (*Fake message*)
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
    83
             TRY (best_tac (!claset addDs [impOfSubs analz_subset_parts,
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
    84
                                           impOfSubs Fake_parts_insert]
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    85
                                    addss (!simpset)) 2)) THEN
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    86
     (*Base case*)
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    87
     fast_tac (!claset addss (!simpset)) 1 THEN
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    88
     ALLGOALS Asm_simp_tac) i;
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    89
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    90
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
    91
(** Theorems of the form X ~: parts (sees lost Spy evs) imply that NOBODY
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    92
    sends messages containing X! **)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    93
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    94
(*Spy never sees another agent's shared key! (unless it's lost at start)*)
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    95
goal thy 
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    96
 "!!evs. evs : otway \
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    97
\        ==> (Key (shrK A) : parts (sees lost Spy evs)) = (A : lost)";
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
    98
by (parts_induct_tac 1);
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
    99
by (Auto_tac());
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   100
qed "Spy_see_shrK";
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   101
Addsimps [Spy_see_shrK];
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   102
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   103
goal thy 
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   104
 "!!evs. evs : otway \
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   105
\        ==> (Key (shrK A) : analz (sees lost Spy evs)) = (A : lost)";
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   106
by (auto_tac(!claset addDs [impOfSubs analz_subset_parts], !simpset));
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   107
qed "Spy_analz_shrK";
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   108
Addsimps [Spy_analz_shrK];
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   109
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   110
goal thy  "!!A. [| Key (shrK A) : parts (sees lost Spy evs);       \
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   111
\                  evs : otway |] ==> A:lost";
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   112
by (fast_tac (!claset addDs [Spy_see_shrK]) 1);
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   113
qed "Spy_see_shrK_D";
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   114
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   115
bind_thm ("Spy_analz_shrK_D", analz_subset_parts RS subsetD RS Spy_see_shrK_D);
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   116
AddSDs [Spy_see_shrK_D, Spy_analz_shrK_D];
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   117
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   118
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   119
(*Nobody can have used non-existent keys!*)
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   120
goal thy "!!evs. evs : otway ==>          \
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   121
\         Key K ~: used evs --> K ~: keysFor (parts (sees lost Spy evs))";
2160
ad4382e546fc Simplified new_keys_not_seen, etc.: replaced the
paulson
parents: 2131
diff changeset
   122
by (parts_induct_tac 1);
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   123
(*Fake*)
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   124
by (best_tac
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   125
      (!claset addIs [impOfSubs analz_subset_parts]
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   126
               addDs [impOfSubs (analz_subset_parts RS keysFor_mono),
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   127
                      impOfSubs (parts_insert_subset_Un RS keysFor_mono)]
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   128
               addss (!simpset)) 1);
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   129
(*OR1-3*)
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   130
by (REPEAT (fast_tac (!claset addss (!simpset)) 1));
2160
ad4382e546fc Simplified new_keys_not_seen, etc.: replaced the
paulson
parents: 2131
diff changeset
   131
qed_spec_mp "new_keys_not_used";
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   132
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   133
bind_thm ("new_keys_not_analzd",
2032
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
   134
          [analz_subset_parts RS keysFor_mono,
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
   135
           new_keys_not_used] MRS contra_subsetD);
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   136
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   137
Addsimps [new_keys_not_used, new_keys_not_analzd];
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   138
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   139
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   140
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   141
(*** Proofs involving analz ***)
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   142
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   143
(*Describes the form of K and NA when the Server sends this message.  Also
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   144
  for Oops case.*)
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   145
goal thy 
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   146
 "!!evs. [| Says Server B                                                 \
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   147
\            {|NA, X, Crypt (shrK B) {|NB, Key K|}|} : set_of_list evs;   \
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   148
\           evs : otway |]                                                \
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   149
\     ==> K ~: range shrK & (EX i. NA = Nonce i) & (EX j. NB = Nonce j)";
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   150
by (etac rev_mp 1);
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   151
by (etac otway.induct 1);
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   152
by (ALLGOALS (fast_tac (!claset addss (!simpset))));
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   153
qed "Says_Server_message_form";
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   154
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   155
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   156
(*For proofs involving analz.*)
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   157
val analz_Fake_tac = 
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   158
    dtac OR2_analz_sees_Spy 4 THEN 
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   159
    dtac OR4_analz_sees_Spy 6 THEN
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   160
    forward_tac [Says_Server_message_form] 7 THEN assume_tac 7 THEN 
2451
ce85a2aafc7a Extensive tidying and simplification, largely stemming from
paulson
parents: 2417
diff changeset
   161
    REPEAT ((eresolve_tac [exE, conjE] ORELSE' hyp_subst_tac) 7);
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   162
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   163
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   164
(****
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   165
 The following is to prove theorems of the form
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   166
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   167
  Key K : analz (insert (Key KAB) (sees lost Spy evs)) ==>
2451
ce85a2aafc7a Extensive tidying and simplification, largely stemming from
paulson
parents: 2417
diff changeset
   168
  Key K : analz (sees lost Spy evs)
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   169
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   170
 A more general formula must be proved inductively.
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   171
****)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   172
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   173
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   174
(** Session keys are not used to encrypt other session keys **)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   175
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   176
(*The equality makes the induction hypothesis easier to apply*)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   177
goal thy  
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   178
 "!!evs. evs : otway ==> \
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   179
\  ALL K KK. KK <= Compl (range shrK) -->                      \
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   180
\            (Key K : analz (Key``KK Un (sees lost Spy evs))) = \
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   181
\            (K : KK | Key K : analz (sees lost Spy evs))";
2032
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
   182
by (etac otway.induct 1);
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   183
by analz_Fake_tac;
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   184
by (REPEAT_FIRST (resolve_tac [allI, impI]));
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   185
by (REPEAT_FIRST (rtac analz_image_freshK_lemma ));
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   186
by (ALLGOALS (asm_simp_tac analz_image_freshK_ss));
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   187
(*Base*)
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   188
by (fast_tac (!claset addIs [image_eqI] addss (!simpset)) 1);
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   189
(*Fake, OR2, OR4*) 
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   190
by (REPEAT (spy_analz_tac 1));
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   191
qed_spec_mp "analz_image_freshK";
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   192
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   193
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   194
goal thy
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   195
 "!!evs. [| evs : otway;  KAB ~: range shrK |] ==>              \
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   196
\        Key K : analz (insert (Key KAB) (sees lost Spy evs)) = \
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   197
\        (K = KAB | Key K : analz (sees lost Spy evs))";
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   198
by (asm_simp_tac (analz_image_freshK_ss addsimps [analz_image_freshK]) 1);
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   199
qed "analz_insert_freshK";
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   200
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   201
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   202
(*** The Key K uniquely identifies the Server's  message. **)
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   203
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   204
goal thy 
2451
ce85a2aafc7a Extensive tidying and simplification, largely stemming from
paulson
parents: 2417
diff changeset
   205
 "!!evs. evs : otway ==>                                                      \
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   206
\   EX B' NA' NB' X'. ALL B NA NB X.                                          \
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   207
\     Says Server B {|NA, X, Crypt (shrK B) {|NB, K|}|} : set_of_list evs --> \
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   208
\     B=B' & NA=NA' & NB=NB' & X=X'";
2032
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
   209
by (etac otway.induct 1);
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   210
by (ALLGOALS (asm_simp_tac (!simpset addsimps [all_conj_distrib])));
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   211
by (Step_tac 1);
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   212
(*Remaining cases: OR3 and OR4*)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   213
by (ex_strip_tac 2);
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   214
by (Fast_tac 2);
2107
23e8f15ec95f The new proof of the lemma for new_nonces_not_seen is faster
paulson
parents: 2052
diff changeset
   215
by (expand_case_tac "K = ?y" 1);
23e8f15ec95f The new proof of the lemma for new_nonces_not_seen is faster
paulson
parents: 2052
diff changeset
   216
by (REPEAT (ares_tac [refl,exI,impI,conjI] 2));
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   217
(*...we assume X is a recent message, and handle this case by contradiction*)
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   218
by (fast_tac (!claset addSEs sees_Spy_partsEs
2032
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
   219
                      delrules [conjI]    (*prevent split-up into 4 subgoals*)
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
   220
                      addss (!simpset addsimps [parts_insertI])) 1);
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   221
val lemma = result();
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   222
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   223
goal thy 
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   224
 "!!evs. [| Says Server B {|NA, X, Crypt (shrK B) {|NB, K|}|}      \
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   225
\            : set_of_list evs;                                    \ 
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   226
\           Says Server B' {|NA',X',Crypt (shrK B') {|NB',K|}|}    \
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   227
\            : set_of_list evs;                                    \
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   228
\           evs : otway |] ==> X=X' & B=B' & NA=NA' & NB=NB'";
2417
95f275c8476e New tactic: prove_unique_tac
paulson
parents: 2375
diff changeset
   229
by (prove_unique_tac lemma 1);
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   230
qed "unique_session_keys";
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   231
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   232
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   233
(*Crucial security property, but not itself enough to guarantee correctness!*)
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   234
goal thy 
2166
d276a806cc10 Tidying up: removing redundant assumptions, etc.
paulson
parents: 2160
diff changeset
   235
 "!!evs. [| A ~: lost;  B ~: lost;  evs : otway |]                    \
d276a806cc10 Tidying up: removing redundant assumptions, etc.
paulson
parents: 2160
diff changeset
   236
\        ==> Says Server B                                            \
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   237
\              {|NA, Crypt (shrK A) {|NA, Key K|},                    \
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   238
\                Crypt (shrK B) {|NB, Key K|}|} : set_of_list evs --> \
2166
d276a806cc10 Tidying up: removing redundant assumptions, etc.
paulson
parents: 2160
diff changeset
   239
\            Says B Spy {|NA, NB, Key K|} ~: set_of_list evs -->      \
d276a806cc10 Tidying up: removing redundant assumptions, etc.
paulson
parents: 2160
diff changeset
   240
\            Key K ~: analz (sees lost Spy evs)";
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   241
by (etac otway.induct 1);
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   242
by analz_Fake_tac;
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   243
by (ALLGOALS
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   244
    (asm_simp_tac (!simpset addcongs [conj_cong] 
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   245
                            addsimps [not_parts_not_analz, analz_insert_freshK]
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   246
                            setloop split_tac [expand_if])));
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   247
(*OR3*)
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   248
by (fast_tac (!claset delrules [impCE]
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   249
                      addSEs sees_Spy_partsEs
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   250
                      addIs [impOfSubs analz_subset_parts]
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   251
                      addss (!simpset addsimps [parts_insert2])) 3);
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   252
(*OR4, OR2, Fake*) 
2375
14539397fc04 Streamlined some proofs
paulson
parents: 2284
diff changeset
   253
by (REPEAT_FIRST spy_analz_tac);
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   254
(*Oops*) (** LEVEL 5 **)
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   255
by (fast_tac (!claset delrules [disjE]
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   256
                      addDs [unique_session_keys] addss (!simpset)) 1);
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   257
val lemma = result() RS mp RS mp RSN(2,rev_notE);
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   258
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   259
goal thy 
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   260
 "!!evs. [| Says Server B                                                \
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   261
\            {|NA, Crypt (shrK A) {|NA, Key K|},                         \
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   262
\                  Crypt (shrK B) {|NB, Key K|}|} : set_of_list evs;     \
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   263
\           Says B Spy {|NA, NB, Key K|} ~: set_of_list evs;             \
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   264
\           A ~: lost;  B ~: lost;  evs : otway |]                \
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   265
\        ==> Key K ~: analz (sees lost Spy evs)";
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   266
by (forward_tac [Says_Server_message_form] 1 THEN assume_tac 1);
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   267
by (fast_tac (!claset addSEs [lemma]) 1);
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   268
qed "Spy_not_see_encrypted_key";
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   269
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   270
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   271
(*** Attempting to prove stronger properties ***)
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   272
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   273
(*Only OR1 can have caused such a part of a message to appear.
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   274
  I'm not sure why A ~= B premise is needed: OtwayRees.ML doesn't need it.
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   275
  Perhaps it's because OR2 has two similar-looking encrypted messages in
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   276
        this version.*)
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   277
goal thy 
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   278
 "!!evs. [| A ~: lost;  A ~= B;  evs : otway |]                \
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   279
\        ==> Crypt (shrK A) {|NA, Agent A, Agent B|}           \
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   280
\             : parts (sees lost Spy evs) -->                  \
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   281
\            Says A B {|NA, Agent A, Agent B,                  \
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   282
\                       Crypt (shrK A) {|NA, Agent A, Agent B|}|}  \
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   283
\             : set_of_list evs";
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   284
by (parts_induct_tac 1);
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   285
by (Fast_tac 1);
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   286
qed_spec_mp "Crypt_imp_OR1";
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   287
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   288
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   289
(*Crucial property: If the encrypted message appears, and A has used NA
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   290
  to start a run, then it originated with the Server!*)
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   291
(*Only it is FALSE.  Somebody could make a fake message to Server
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   292
          substituting some other nonce NA' for NB.*)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   293
goal thy 
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   294
 "!!evs. [| A ~: lost;  A ~= Spy;  evs : otway |]                        \
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   295
\        ==> Crypt (shrK A) {|NA, Key K|} : parts (sees lost Spy evs) --> \
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   296
\            Says A B {|NA, Agent A, Agent B,                  \
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   297
\                       Crypt (shrK A) {|NA, Agent A, Agent B|}|}  \
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   298
\             : set_of_list evs -->                            \
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   299
\            (EX B NB. Says Server B                           \
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   300
\                 {|NA,                                        \
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   301
\                   Crypt (shrK A) {|NA, Key K|},              \
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   302
\                   Crypt (shrK B) {|NB, Key K|}|}             \
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   303
\                   : set_of_list evs)";
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   304
by (parts_induct_tac 1);
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   305
(*OR1: it cannot be a new Nonce, contradiction.*)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   306
by (fast_tac (!claset addSIs [parts_insertI]
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   307
                      addSEs sees_Spy_partsEs
2032
1bbf1bdcaf56 Introduction of "lost" argument
paulson
parents: 2002
diff changeset
   308
                      addss (!simpset)) 1);
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   309
(*OR4*)
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   310
by (REPEAT (Safe_step_tac 2));
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   311
by (REPEAT (best_tac (!claset addSDs [parts_cut]) 3));
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   312
by (fast_tac (!claset addSIs [Crypt_imp_OR1]
2516
4d68fbe6378b Now with Andy Gordon's treatment of freshness to replace newN/K
paulson
parents: 2451
diff changeset
   313
                      addEs  sees_Spy_partsEs) 2);
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   314
(*OR3*)  (** LEVEL 5 **)
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   315
by (ALLGOALS (asm_simp_tac (!simpset addsimps [ex_disj_distrib])));
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   316
by (step_tac (!claset delrules [disjCI, impCE]) 1);
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   317
(*The hypotheses at this point suggest an attack in which nonce NA is used
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   318
  in two different roles:
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   319
          Says B' Server
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   320
           {|Nonce NAa, Agent Aa, Agent A,
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   321
             Crypt (shrK Aa) {|Nonce NAa, Agent Aa, Agent A|}, Nonce NA,
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   322
             Crypt (shrK A) {|Nonce NAa, Agent Aa, Agent A|}|}
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   323
          : set_of_list evsa;
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   324
          Says A B
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   325
           {|Nonce NA, Agent A, Agent B,
2284
80ebd1a213fd Swapped arguments of Crypt (for clarity and because it is conventional)
paulson
parents: 2264
diff changeset
   326
             Crypt (shrK A) {|Nonce NA, Agent A, Agent B|}|}
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   327
          : set_of_list evsa 
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   328
*)
2131
3106a99d30a5 Changing from the Reveal to the Oops rule
paulson
parents: 2107
diff changeset
   329
writeln "GIVE UP! on NA_Crypt_imp_Server_msg";
2002
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   330
ed423882c6a9 Bad version of Otway-Rees and the new attack on it
paulson
parents:
diff changeset
   331
2052
d9f7f4b2613e Working again with new theory Shared
paulson
parents: 2032
diff changeset
   332
(*Thus the key property A_can_trust probably fails too.*)